Urbanity mapping reveals the complexity, diffuseness, diversity, and connectivity of urbanized areas

Abstract

There are urgent calls for new approaches to map the global urban conditions of complexity, diffuseness, diversity, and connectivity. However, existing methods mostly focus on mapping urbanized areas as bio physical entities. Here, based on the continuum of urbanity framework, we developed an approach for cross-scale urbanity mapping from town to city and urban megaregion with different spatial resolutions using the Google Earth Engine. This approach was developed based on multi-source remote sensing data, Points of Interest – Open Street Map (POIs-OSM) big data, and the random forest regression model. This approach is scale-independent and revealed significant spatial variations in urbanity, underscoring differences in urbanization patterns across megaregions and between urban and rural areas. Urbanity was observed transcending traditional urban boundaries, diffusing into rural settlements within non-urban locales. The finding of urbanity in rural communities far from urban areas challenges the gradient theory of urban-rural development and distribution. By mapping livelihoods, lifestyles, and connectivity simultaneously, urbanity maps present a more comprehensive characterization of the complexity, diffuseness, diversity, and connectivity of urbanized areas than that by land cover or population density alone. It helps enhance the understanding of urbanization beyond biophysical form. This approach can provide a multifaceted understanding of urbanization, and thereby insights on urban and regional sustainability.

Keywords

• Continuum of Urbanity
• Big data
• Mapping
• Spatial regression
• Multiscale